http://arxiv.org/abs/1510.05478
We present 3D hydrodynamic simulations of the adiabatic interaction of a shock with a dense, spherical cloud. We compare how the nature of the interaction changes with the Mach number of the shock, M, and the density contrast of the cloud, chi. We also examine the differences with 2D axisymmetric calculations, perform detailed resolution tests, and compare “inviscid” results to those obtained with the inclusion of a k-epsilon subgrid turbulence model.
We find that resolutions of 32-64 cells per cloud radius are the minimum necessary to capture the dominant dynamical processes in 3D simulations. In contrast to our earlier 2D work, we find that 3D inviscid and k-epsilon simulations typically show very good agreement. As such, there does not appear to be any compelling reason for using the k-epsilon subgrid model in 3D calculations, though it remains very useful for 2D calculations. Clouds accelerate and mix up to 5 times faster when they are poorly resolved. This has implications for numerical simulations of multi-phase flows where a fast, low density medium interacts with slower, higher density clouds (e.g., galactic winds).
The interaction proceeds very similarly in 2D and 3D – although non-azimuthal modes lead to different behaviour, there is very little effect on key global quantities such as the lifetime of the cloud and its acceleration. We do not find significant differences in the hollowing or “voiding” of the cloud between 2D and 3D simulations with M=10 and chi=10, in contradiction to expectations. This may be due to the softer edges used for our clouds.
The biggest differences between our 2D and 3D calculations are found when M=1.5 and chi=10 – the cloud is destroyed more rapidly in 2D simulations, perhaps because secondary vortices form earlier and are more prevelant in the higher resolution 2D simulations.
J. Pittard and E. Parkin
Tue, 20 Oct 15
17/92
Comments: 32 pages, 35 figures, submitted to MNRAS
You must be logged in to post a comment.